CSA Social Marketing: Music Clusters

Genre level correlations

“Which genres are alike in terms of their audiences?””

In the genre × genre matrix, similarity values capture shared audience bases between genres. Use genre × genre when you care about brand/genre positioning, cross-genre affinities, playlisting, or recommender systems.

Correlation Matrix

Interactive Correlation Matrix

Optimal Number of Clusters

Hierarchical Clustering - K = 4

Hierarchical Clustering - K = 5

Clustered Heatmap

Interactive Clustered Heatmap

K Means - K = 4

K Means - K = 5

Circlize

Network

Network 2

Network 3

Person level correlation

“Which people are alike in their genre preferences?”

In the person × person matrix, similarity values capture shared taste patterns between individuals. Use person × person when you care about segmentation, consumer clusters, marketing personas.

3 cluster solution

Cluster interpretation

• 1: Mainstream Western Pop/Rock
• 2: Mainstream Afro-diasporic
• 3: Niche/Eclectic/Experimental

4 cluster solution

Cluster interpretation

• 1: Mainstream Western Pop/Rock
• 2: Niche/Eclectic/Experimental
• 3: Mainstream Afro-diasporic
• 4: Non-Mainstream Rock (Headbangers)